"Leader-Follower" Singing in Red-Winged Blackbirds

Abstract

The males of many songbird species sing speciestypical songs at characteristic rates throughout the breeding season. These songs are correlated with establishment and maintenance of territory boundaries as well as with mate acquisition. Many male songbirds will sing regardless of whether or not conspecific males are nearby. However, the rate of song delivery is often strongly influenced by the presence of both conspecific males and females (Smith et al., unpubl. data; Dinsmore, Auk 86:139-140, 1969). Red-winged Blackbirds usually nest in colonies, occupying contiguous territories. During a study of this species, we noticed that when a potentially intruding male sang, he was often answered by the resident male. We termed this phenomenon leaderfollower the threatening male being the leader (A) and the resident, the follower (B). It usually occurred at disputed boundaries between territories. Throughout the breeding season, males sing short, stereotyped, conc-a-ree songs coupled with a visual display 1-9 times per minute. Males are often involved in conflicts with one another early in the breeding season when boundaries are being disputed. The data we discuss here were collected from a population of marked birds which we observed during the 1976 and 1977 breeding seasons, in Old Field, New York. The interval between the songs of two male blackbirds sometimes varies predictably, in which case it may indicate to a threatening male that a particular resident male is paying attention to him. A leader-follower singing relationship developed wherever a territorial boundary was threatened by a non-resident or by a neighboring territorial male. In order to standardize data collection, we defined a boundary encounter as a situation in which singing males were 10 m or less apart, separated by a boundary. To quantify this apparent interactive singing, we used a tone encoder system and noted the song delivery of the males and their territorial status. We plotted as a histogram the intervals between songs of males A and B (A -) B) and those intervals between B's song and the following song of A (B -> A). We predicted that if the songs were sung independently of one another, the intervals would be highly variable. Furthermore, this variability should equal that of birds singing at distances greater than 10 m apart. In order to test this hypothesis, we computed a set of intersong intervals using the mean and variance of intervals between songs of each bird during an interactive singing period. We then examined the resulting distribution of the intervals, first assuming that the birds were singing independently. We measured intersong intervals and calculated means and standard deviations of these intervals for each male. We then computed two independent random song delivery sequences and plotted them onto the same time axis in order to compare the series of times. The computed A -) B and B -) A intervals were analyzed (Fig. 1, upper graph) and compared with the observed intervals (Fig. 1, lower graph). In the example shown, our computed interval histogram shows that the distributions of the A -> B and B -A intervals are not significantly different from one another (t-test, two-tailed, P > 0.3). The interval histogram for the observed data, however, differs from that of the computed independent distributions. First, the